Beth Adelson

Problem solving and the development of abstract categories in programming languages

The question of how novice and expert computer programmers represent and use programming concepts is addressed here. Lines of programming code forming three complete programs were presented one at a time and in random order in a multitrial free recall procedure. Qualitative and quantitative measures revealed clear but different subjective organization in the two groups. The novices used a syntax-based organization, whereas the experts used a more abstract hierarchical organization based on principles of program function.

A model of software design

In this article we present a preliminary cognitive model of the process of software design. Our goal was to develop a model of expert problem‐solving skills for a task in which domain knowledge played an extensive role. In our model the process of design is captured via goals‐and‐operators interacting with a knowledge base. We have defined the goals and operators as ones which are general to design, rather than specific to the current task. In addition, we have structured the atomic level operators so that they are able to access domain specific knowledge acquired through experience. This structure enables both general processes and domain specific knowledge to play critical roles in producing any particular design artifact. From our protocol analysis we have built a model which unites several recurring behaviors into an interpretable whole. the behaviors we account for include the building of mental models, mental simulation, and balanced development.

Semantic specificity in cued recall

In three cued recall experiments, extralist retrieval cues that were congruous in meaning to an encoded target pair of words produced better recall of the targets than wh6n the cuetarget relation was incongruous. However, this result, which differs from that of other experiments, depended in some cases on scoring recall of a target pair when either member of the pair was recalled. It is argued that (1) pairs of words are typically stored as higher order units, (2) the best test procedure is to request recall of both members of the pair when an extralist cue is presented, and (3) semantic features provide an important dimension in the mnemonic representation of word events.

Comparing Natural and Abstract Categories: A Case Study from Computer Science

Are abstract and concrete objects categorized similarly? In studies of psychological categories of concrete objects, Rosch and her colleagues have documented two phenomena (Rosch, Mervis, Gray, Johnson, & BoyesBraem, 1976). They found a basic level of conceptual abstraction and a perception of categorical structure based on family resemblance. To parallel Rosch’s experiments, this article reports three studies of categories of abstract objects. Working within a Roschian experimental paradigm, we ask whether computer scientists identify a natural, basic level of abstraction similar to that found for concrete objects? We also ask whether these categories exhibit prototypicality and family resemblance characteristics? The question of whether abstract and concrete objects are similarly categorized is motivated by the results of Adelson (1981) which suggested that expert programmers chunk single lines of code into conceptual objects. Adelson asked novice and expert programmers to recall a set of 16 lines of programming code which had been presented in random order. Although

Evocative agents and multi-media interface design

This paper describes research which focusses on the issue of possible roles for computerized agents within multimedia educational software.

Cognitive research: Uncovering how designers design; cognitive modeling: Explaining and predicting how designers design

In this paper we describe: 1) A research approach that can produce cognitive models of the design process and 2) A particular model that illustrates the approach.Cognitive models describe the complex behaviors required to solve engineering problems in terms of underlying functional mechanisms. Because the models describecognitive systems at the level of functional mechanisms, they can generate explanations and predictions about the behaviors being studied. This makes the model useful to theorists; it also provides insights concerning support and teaching tools for practitioners.To illustrate our approach we present a model of learning to program by analogy. Within the model we describe mechanisms that simplify the processes of mapping, evaluating and debugging newly acquired knowledge in a complex domain. We also discuss the theoretical and practical implications of the model.

Issues in scientific creativity: insight, perseverance and personal technique: Profiles of the 2002 Franklin Institute Laureates

Abstract In this paper, we reflect on some issues basic to the understanding of scientific creativity, including the nature of the “Ah Hah” experience; the relationship of that experience to perseverance; the development of a technique that fits ones personal strengths; the tension between building on and breaking out of intellectual traditions; and, because science is a social endeavor, some ways in which truly novel work has gained acceptance. This paper grows out of a set of interviews with the 2002 Franklin Institute Laureates. The interviews lead us to pay attention to the pairing of analytic ability with synthetic ability. The literature on problem solving has emphasized the importance of synthetic ability in scientific creativity, but has not paid equal attention to this pairing. Looking just at analytic ability, we find the Laureates continually rely on this ability when confronting new realms. Thus, instead of having felt anxiety as they moved into the fields which they have opened up, they experienced emotions ranging from comfort to joy. The Laureates often move into new fields. Their analytic skill allows them to do so with confidence and their creative curiosity gives them a taste for it. And working in their new and “uncrowded” fields they are free to pursue their own take on their own problems. Perhaps this lack of competitive pressure contributes to the personal satisfaction that runs through this group.

Modeling Software Design Within a Problem-Space Architecture

In this paper1 we describe research on modeling software design skills within the Soar problem-solving architecture (Laird, Newell & Rosenbloom, 1987; Steier & Newell, 1987). We focus on an analysis of expert designers of communications systems designing an electronic mail system.

PSYCHOLOGICAL SCIENCE AND ANALOGICAL REMINDING IN THE DESIGN OF ARTIFACTS

In this article, we discuss two approaches to design currently prevalent in the field of human- computer interaction. Proponents of one approach advocate working from first principles of psychological science; practitioners of the second, engineering, approach find that successful design is often the result of analogical discovery. However, these approaches have not yet been evaluated for the usability of their methods. Can designers work in each of these ways? What are the difficulties? How can they be overcome? In this paper, we take a first step in addressing these questions. We analyze the form in which principle-based design poses its questions, thereby uncovering some of the difficulty that is encountered in this way of working. Furthermore, we analyze the form in which engineering problems are expressed and identify an advantage of this typically analogical way of working. We then suggest a way in which analogy-based design can supplement both principle-based and empirically driven design.

Constructs and phenomena common to the semantically-rich domains

This article looks at some phenomena that recur during expert problem solving across domains including computer science, physics, chess, and mathematics. Several of the mechanisms which have been hypothesized to underlie performance in one domain are then used to explain the results found in the others. Competing explanations, mechanisms which are not fully specified, and untested empirical issues are also discussed.